Background Factors influencing choice and use of contraceptives among HIV discordant couples have important policy and programmatic implications for HIV and pregnancy prevention. We investigated factors among women and men in an HIV serodiscordant couple enrolled in a phase 3 clinical trial of Partners PrEP study.

Methods Prospective sub-study enrolled 385 men and women selected from 629 couples. Family planning counselling and contraceptives were offered to participants at each visit. Questionnaires were administered at enrolment and after 6 months and 12 months to document contraceptive knowledge, preference and use. A logistic generalised estimating equation was used. IRB reviewed written consents.

Results The average age was 28.5 years (SD = 8.5) for female and 34.9(SD = 10.3) years for males. The median number of children was 1(IQR = 0–3). At baseline, knowledge of contraceptives was 99% for both men and women, and 143 (37%) reported current use of any contraceptive (male condoms = 50%, injectables = 31%, implants = 9%, oral pills = 4%, other = 6%). Of the 242 not using contraceptives at baseline, 182 (75%) wanted to use a contraceptive; 98 (54%) of them specified their preferences including 52% injectables, 22% male condoms, 11% oral pills, 9% implants and 6% other. At 6 month follow-up 302 (78%) and 12-month follow-up visit 274 (71%) reported use of contraception; an increased odds (OR = 1.74, 95% CI = 1.64–3.53; OR = 1.74, 95% CI = 1.20–2.53, respectively) of contraceptive use over baseline HIV-negative participants were more likely to use contraception (including condoms) than HIV-positive participants (OR = 1.49.95% CI (1.05–2.01). Odds of contraceptive use rose with the number of children one had (OR = 1.09, 95% CI 1.00–1.19). Main reason for not using contraceptives during follow up was fear of side effects (OR = 0.49, 95% CI 0.44–0.55).

Conclusion Education and counselling provided at this site led to an almost doubling of contraceptive usage among HIV serodiscordant couples. Policymakers can adopt this to address gap between knowledge and usage.